OpenAI confirmed a significant global outage affecting its ChatGPT and Codex services on April 20, 2026, leaving thousands of developers, professionals, and students unable to access the company’s primary artificial intelligence tools. The disruption, which was first acknowledged by the company at 10:35 a.m. ET, has impacted multiple components of the OpenAI ecosystem, including the web interface, mobile application, and API platform.
According to data from the service monitoring site Downdetector, reports of issues began to spike around 10:05 a.m. ET. The outage appeared to have a particularly heavy impact in the United Kingdom, where reports peaked at over 8,000. In the United States, Downdetector recorded more than 2,700 reports by 10:52 a.m. ET. Users in India also reported significant disruptions, with over 3,000 complaints registered. Statistical analysis of these reports indicated that 86 percent of users were unable to access core ChatGPT functionalities, while others cited specific difficulties with logins and general application stability.
Technical failures reported by users included persistent 403 Forbidden errors, gateway timeouts, and an inability to load existing chat histories. OpenAI’s official status page confirmed that 12 distinct components of the ChatGPT system and five components of the Codex system were affected. Impacted features included voice mode, search capabilities, file uploads, and image generation. Furthermore, the outage extended to the API Platform, disrupting third-party applications and enterprise workflows that rely on OpenAI’s infrastructure.
OpenAI issued a statement confirming the investigation into the degraded performance and noted that users were currently unable to load ChatGPT and Codex. The company specifically highlighted an issue affecting ChatGPT Business users, where those attempting to upgrade accounts or add new seats encountered errors that could persist for up to an hour. While the desktop version of the platform remained largely non-functional for the duration of the peak outage, some users reported that the mobile application remained partially operational. OpenAI officially advised users to utilize the mobile interface as a temporary workaround while engineering teams worked to stabilize the primary desktop services.
As of 11:13 a.m. ET, OpenAI provided a status update indicating that they had applied a mitigation strategy and were monitoring the recovery process. However, many users continued to report intermittent connectivity issues and authentication failure messages. This event follows a pattern of occasional infrastructure challenges for the company as it scales to meet the demands of a user base that now numbers in the hundreds of millions. No official root cause for the April 20 incident has been disclosed by OpenAI at this time.